4 ms·
Greed is good as Gordon Gecko put it, and science concurs. http://time.com/41680/greed-is-good-science-proves/ http://time.com/41680/greed-is-good-science-prove
by draugadrotten 11y ago
Greed is good as Gordon Gecko put it, and science concurs. http://time.com/41680/greed-is-good-science-proves/ http://time.com/41680/greed-is-good-science-proves/
- Intermernet 11y agoThat Time article is linked on a bunch of fora, but none of them have a link (or reference) to the actual study. Any idea if it's available? I suspect that Time's treatment of it is not as nuanced as the original authors would have liked, and even then, the Time article doesn't actually concur that "Greed is good". They just have that phrase in the headline.
- rgbrenner 11y agoI think this is the paper: http://www.nature.com/ncomms/2014/140326/ncomms4526/full/ncomms4526.html http://www.nature.com/ncomms/2014/140326/ncomms4526/full/nco...
- Intermernet 11y agoThanks, much obliged. I wish articles would provide proper references!
- mikeash 11y agoIf they did that, then more people would read the actual study and realize that the journalistic summary is completely full of it. Can't have that.
- crpatino 11y agoNo, that's what the title says, but it is click bait. What the actual article says is that in social groups with well defined hierarchies, higher ranking members tend to contribute more than average to the common good (though this might be biased due to their role in defense against external threats). This old fashioned notion that the rich are expected to hit above their weight level is called Noblesse Obligue, and it's almost the exact opposite of Gecko's doctrine. What reasons there might be to have come up with such a title/conclusion given the information available? I can only think of one: Assholeness apologetics.
- ericdykstra 11y agoGreedy algorithms come up with simple, but often suboptimal decisions when the whole picture is taken into account.
- dllthomas 11y agoAs opposed to people, who generally come up with decisions of varying degrees of complexity, which are still suboptimal when the whole picture is taken into account.
- aikah 11y agoDefine "good".
- idibidiart 11y agoIf you'd like someone to defined Good for you, I'm afraid putting it in quotes, like "stupid", is not a good start.
- uhwhat 11y ago*Scientists interpreted observation as clickbait worthy title for blog posts